Lipschitz Lifelong Reinforcement Learning
This work addresses the challenge of efficient knowledge transfer across sequential tasks in reinforcement learning, which is incremental but provides theoretical guarantees.
The authors tackled the problem of knowledge transfer in lifelong reinforcement learning by introducing a novel metric between Markov Decision Processes and proving that close MDPs have close optimal value functions, leading to a value-transfer method with improved convergence rates and no negative transfer with high probability.
We consider the problem of knowledge transfer when an agent is facing a series of Reinforcement Learning (RL) tasks. We introduce a novel metric between Markov Decision Processes (MDPs) and establish that close MDPs have close optimal value functions. Formally, the optimal value functions are Lipschitz continuous with respect to the tasks space. These theoretical results lead us to a value-transfer method for Lifelong RL, which we use to build a PAC-MDP algorithm with improved convergence rate. Further, we show the method to experience no negative transfer with high probability. We illustrate the benefits of the method in Lifelong RL experiments.